Notice how the author assumes UUID v4[1] in the conversation. There are very few reasons to use the other versions but we are still paying for their price in code complexity all the time.
Look at this UUID parsing code: https://github.com/sporkmonger/uuidtools/blob/master/lib/uui...
What it really should be is `[uuid_string.gsub('-', '')].pack('H*')` (for non-rubyists: remove the dashes, decode the hex back to binary).
Their representation is also not that good since hex encoding is not very compact.
I guess what I'm trying to say is that UUIDs are often used as a default unique identifiers but they are actually not that good.
[1]: https://en.wikipedia.org/wiki/Universally_unique_identifier#...
Notice how the author assumes UUID v4[1] in the conversation. There are very few reasons to use the other versions but we are still paying for their price in code complexity all the time.
Look at this UUID parsing code: https://github.com/sporkmonger/uuidtools/blob/master/lib/uui...
What it really should be is `[uuid_string.gsub('-', '')].pack('H*')` (for non-rubyists: remove the dashes, decode the hex back to binary).
Their representation is also not that good since hex encoding is not very compact.
I guess what I'm trying to say is that UUIDs are often used as a default unique identifiers but they are actually not that good.
[1]: https://en.wikipedia.org/wiki/Universally_unique_identifier#...